每日同步 Arena 公开数据
AI 模型排行榜
从匿名对战投票与真实任务反馈，了解 AI 模型的表现。覆盖文本、编程、视觉、图像和视频 12 个榜单。
- 最近更新
- 2026-10-12
- 上榜模型
- 712
- 榜单
- 12
- 怎么排的
各榜前三
点任意榜单查看完整排名
Agent
真实任务中的执行、工具使用与用户反馈
1
Claude Opus 5.5 (Max)17.89%±4.24%
Claude Opus 5.5 (Max)17.89%±4.24%2
GPT 6.1 Sol (Max)13.81%±2.34%
GPT 6.1 Sol (Max)13.81%±2.34%3GPT 6 Sol (Max)12.60%±2.60%
GPT 6 Sol (Max)12.60%±2.60%文本
对话、写作、推理等文字任务
1
gemini-4-argon-high1,525±9
gemini-4-argon-high1,525±92claude-opus-5.5-high1,507±8
claude-opus-5.5-high1,507±83claude-opus-4-6-high1,504±3
claude-opus-4-6-high1,504±3网页开发
从需求到网页的代码生成能力
1claude-opus-5.5-max1,813+15/-15
claude-opus-5.5-max1,813+15/-152gpt-6-astra-max1,786+11/-11
gpt-6-astra-max1,786+11/-113claude-sonnet-5.5-xhigh1,774+15/-15
claude-sonnet-5.5-xhigh1,774+15/-15视觉
图片、图表与界面的理解能力
1claude-fable-5-high1,308±7
claude-fable-5-high1,308±72
qwen3.8-max1,301±7
qwen3.8-max1,301±73claude-opus-4-6-high1,299±6
claude-opus-4-6-high1,299±6文档
长文档阅读、理解与问答
1claude-opus-5-high1,516±7
claude-opus-5-high1,516±72claude-fable-5.1-max1,513±15
claude-fable-5.1-max1,513±153claude-opus-4-6-high1,507±6
claude-opus-4-6-high1,507±6搜索
联网检索与信息整合
1gpt-5.6-sol-xhigh1,257±7
gpt-5.6-sol-xhigh1,257±72claude-opus-4-6-search1,253±5
claude-opus-4-6-search1,253±53gpt-5.5-search1,242±5
gpt-5.5-search1,242±5图生网页
从设计图片生成网页代码
1claude-opus-5.5-max1,749+30/-30
claude-opus-5.5-max1,749+30/-302claude-sonnet-5.5-xhigh1,740+20/-20
claude-sonnet-5.5-xhigh1,740+20/-203gpt-6-astra-max1,731+13/-13
gpt-6-astra-max1,731+13/-13文生图
根据文字描述生成图像
1gpt-image-2.5-sunburst1,425±7
gpt-image-2.5-sunburst1,425±72gpt-image-2.5-flare1,398±7
gpt-image-2.5-flare1,398±73gpt-image-2 (medium)1,383±4
gpt-image-2 (medium)1,383±4图片编辑
按指令调整和修改图片
1gpt-image-2.5-sunburst1,524±5
gpt-image-2.5-sunburst1,524±52gpt-image-2.5-flare1,481±5
gpt-image-2.5-flare1,481±53gpt-image-2 (medium)1,462±3
gpt-image-2 (medium)1,462±3文生视频
根据文字描述生成视频
1gemini-omni-1.1-flash1,516±15
gemini-omni-1.1-flash1,516±152gemini-omni-flash1,513±9
gemini-omni-flash1,513±93
flux-3-video1,493±17
flux-3-video1,493±17图生视频
将静态图片转为动态视频
1
minimax-h31,495±5
minimax-h31,495±52gemini-omni-1.1-flash1,488±11
gemini-omni-1.1-flash1,488±113wan3.01,480±11
wan3.01,480±11视频编辑
按指令修改视频内容
1wan3.01,414±26
wan3.01,414±262
dreamina-seedance-2.5-720p1,410±26
dreamina-seedance-2.5-720p1,410±263minimax-h31,392±19
minimax-h31,392±19